Each year, at least 1.4 million people sustain TBI, with over 1.1 million treated and released from the emergency department (ED). Approximately 125,000 patients with TBI, typically those more severely injured, experience permanent disability as a result of damage to the brain. On the other hand, mild TBI, accounting for at least 75% of all TBI, results in more subtle functional and cognitive deficits that often go undetected in the acute setting. These patients can experience drastic changes in their quality of life, have difficulties returning to daily activities and may be unable to return to work for weeks or months. It is estimated that mild TBI alone costs the United States in excess of $17 billon per year in long-term care and lost productivity. Lack of objective criteria for diagnosing and classifying acute TBI presents a significant impediment to developing therapies and to providing uniform quality of care. Patients with a history of head injury often receive a screening brain scan using computed tomography (CT) in the ED despite a low sensitivity for detecting mild injury. The majority of patients discharged directly from the ED are assigned the vague diagnosis of concussion and left to their own accord to seek further medical support, despite a persistent and complex post concussive syndrome that includes complaints of headaches, dizziness, impaired attention, trouble concentrating, poor memory, and emotional problems. The objective of this project is to generate natural history data for cohort-based comparisons to serve as the basis for future hypothesis-driven protocols and to contribute to the clinical and physiological understanding of traumatic brain injury (TBI) through the description of manifestations of the injury and the relationship among radiological, hematological, clinical variables and standard functional/cognitive outcome measures. This is a prospective cohort study of subjects with known and suspected non-penetrating acute traumatic brain injury. Subjects presenting to the emergency department or trauma service at participating hospitals with a history of recent head injury are studied during the course of their hospital stay and after discharge using radiological, hematological, clinical and functional/cognitive outcome measures. Subjects will be stratified according to findings into cohorts for comparison. The design is intentionally broad in scope to allow acquisition of initial data for the development of future hypothesis-driven protocols. Research performed under this protocol will not interfere with standard of care and subjects will not be treated with experimental therapies as part of the research study. Data collected under this research study may be shared without personal identifiers with other researchers if subjects approve this option on the informed consent. To date, more than 1,000 adult subjects with history of recent head injury with or suspected non-penetrating acute TBI will be enrolled. Subjects having varying degrees of TBI severity will be recruited from the collaborative programs between NIH and non-NIH hospitals. Approximately 80% of subjects are classified as mild TBI, concussion, or no injury, with approximately one half of subjects enrolled being discharged directly from the emergency department. Subjects are offered the option of coming to the NIH clinical center to be imaged at very high fields (7T), imaged with positron emission tomography (PET), or followed for up to 5 years under a collaborative project.